9th Nov 2021
Mr Cheng Wai Keung, Chairman, SingHealth
Ms Yong Ying-I, Chairman, SGInnovate, and Permanent Secretary for Ministry of Communications and Information
Professor Ivy Ng, Group CEO, SingHealth
Dr Lim Jui, CEO, SGInnovate
Distinguished guests
Excellencies
Ladies and gentlemen
Introduction
1. I am very pleased to join everyone, so early in the morning to witness the signing of the Memorandum of Understanding (MOU) between SingHealth and SGInnovate. We are able to have this event online amidst the ongoing COVID-19 pandemic, due to technology. But actually, the rules do allow us to meet physically without food and with our masks on, the problem is we don’t get to see each other’s faces. So it is always a trade-off now. But no doubt, the world has changed drastically in the last 20 years. We have information and vast computing power literally in our hands. And Artificial Intelligence (AI) tells us where to go, what to buy and what to watch. We are witnessing the start of a revolution in healthcare too, starting with smart devices tracking our vital signs, our activities and recommending customised workout plans, habits, that benefit us in the best possible way.
AI is Already a Part of our Lives
2. Three inter-connected factors made all these possible.
3. First, the exponential rise in the availability of data – the majority coming from the Internet. And these are mostly user-generated raw data from social media and the “Internet of Things”, which are captured by computers to learn the underlying trends and patterns from a wide range of cases and examples. It is as if the computer had walked the many journeys undertaken by many people, and extracted their habits, their lifestyle, their wisdom in a focused topic.
4. Second, is machine learning, in particular deep learning, which by processing that vast amount of data, mimic the human’s ability to sense, comprehend and learn continuously. It is like the craftsmen of old, who hone their skills over a lifetime, except the machine does so in a compressed time when fed with a treasure trove of data.
5. Third, the ever-faster processing speed and ever-growing on-board memory capacity of computers, which continues to validate Moore’s Law.
6. The result is AI that is changing the world as we know it. The seminal breakthroughs in AI technology were reported in rather understated way, in its ability to outplay humans in certain games.
7. For example, in 1997, Deep Blue beat a reigning world chess champion under tournament conditions. But Deep Blue essentially memorised all possible moves and countermoves and demonstrated how much and how fast computers can store and process information. That was what the victory meant.
8. And that computer ability was further demonstrated in 2011, when Watson beat two of the greatest human champions of the Jeopardy game show. The knowledge of computer demonstrated then had gone beyond chess moves, to all facets of human knowledge.
9. And then in 2016, AlphaGo defeated the top human players of the game of Go, a far more sophisticated game of strategy with, I believe, infinite moves. This is AI and machine now learning in action, not just memorising and storing moves.
10. To beat humans in our own games are bell-weather events that herald the possibility of AI. It is a signal to humankind, that AI can potentially transform life as we know it. We are now faced with the choice of riding the AI wave to improve lives and augment our relevance, or be made redundant by it.
11. So today, let me speak about the relevance of AI in just one sector, which is healthcare.
AI in Healthcare
12. Healthcare is a beneficiary of AI. Computer scientists have successfully cracked a decades-old protein-folding challenge with an AI system that could eventually help identify new treatments for diseases.
13. Locally, scientists used the capability of AI in understanding natural language, as shown in the Jeopardy game, to develop chatbots. A useful chatbot is Doctor COVID, which can speak to the migrant workers infected with COVID-19 in their languages, and enables our healthcare teams to better monitor their clinical and mental wellness at community care facilities. And this was mentioned by Professor Ivy Ng just now.
14. Changi General Hospital and the Integrated Health Information Systems (IHiS) also co-developed a deep-learning system to estimate the likelihood that a patient may develop severe pneumonia based on an automated assessment of chest X-rays and this allows early intervention from doctors.
15. A joint research team from the Singapore National Eye Centre, Singapore Eye Research Institute and NUS School of Computing, they developed what is called SELENA+, which stands for Singapore Eye LEsioN Analyser Plus. It is an AI system which uses retina images to detect major eye diseases. It has been deployed to 20 polyclinics. SELENA+ will help us overcome our shortage of trained healthcare professionals who can detect high-risk patients early and then prevent future blindness.
16. Now, using wearable sensors and cloud computing, patients’ real-time vital sign readings and symptoms can be remotely monitored. The data is then analysed using AI to predict early complications of diseases.
17. So beyond all these examples of current applications of AI in healthcare, there are also immense opportunities to leverage AI to improve care, to advance medical science, and ultimately save more lives.
18. Let me briefly mention three promising initiatives, not necessarily technological in nature, to boost the development of healthcare in Singapore, and unleash the potential of it.
Initiatives to boost the development of Health AI
19. First, building up our healthcare database, as the foundation to drive AI innovation. And this is a key reason for developing the National Electronic Health Record (NEHR) in 2011. NEHR brought together clinical summary records from different care providers, mostly from public healthcare institutions, but increasingly also private providers.
20. Beyond clinical records, we are generating genomic information of 100,000 consenting individuals from Singapore. We are accumulating anonymised lifestyle data for research through various national programmes and partnerships with the private sector, such as the National Steps Challenge and LumiHealth (and this is HPB’s collaboration with Apple). These data will tune the AI systems to our population, and our local context, so as to better predict health risks and outcomes, and improve recommendations on interventions that could slow down or reverse disease progression.
21. Second initiative, we have put in place regulations and guidelines that encourage and enable the responsible use of AI in healthcare. MOH, the Health Sciences Authority (HSA) and IHiS have codified good practices through the AIHGlE (Artificial Intelligence in Healthcare Guidelines) project, and this support the safe growth of AI in healthcare. This will enhance trust in AI’s legitimate and broad usage in healthcare.
22. Third initiative, we are encouraging multidisciplinary collaborations and innovations in AI projects. The most impactful innovations involve the combination of skills and talents across many domains and disciplines, technical and non-technical. We must therefore tap on the range of capabilities in our Research, Innovation and Enterprise (RIE) ecosystem, in private as well as public sectors.
23. One good example is the AI in Health Grand Challenge run by AI Singapore. With funding from the National Research Foundation, AI Singapore has allocated up to S$35 million to support the co-development of AI solutions between public healthcare institutions and institutes of higher learning. The aim is to enhance primary care, health promotion, wellness maintenance, and disease management.
Partnership between SingHealth and SGInnovate
24. Which brings us to today’s event. The MOU that will be signed today marks a new milestone towards a more integrated and forward-looking partnership between SingHealth and SGInnovate. SGInnovate’s capabilities are to translate scientific research into ‘technology intensive’ applications. SingHealth, on the other hand, is home to our oldest healthcare institution, the Singapore General Hospital, celebrating its bicentenary this year, and one of the most renowned in the world for delivering excellent care to its patients.
25. So together, they can advance health AI innovations in meaningful ways, building upon the national foundations we have set and will continue to build, that I mentioned earlier.
26. SingHealth and SGInnovate will organise knowledge sharing events, and build up Communities of Practice, to bring together deep expertise in healthcare and AI, including the Deep Tech community of corporates, start-ups and innovators.
27. This partnership will also focus on providing start-ups with the resources they need to develop and grow. This includes connecting innovators to potential investors, and supporting start-ups in developing business plans, go-to-market strategies, and commercial networking.
28. SGInnovate will also encourage Deep Tech entrepreneurs to work alongside clinician scientists, clinician investigators, doctors, nurses, and allied health professionals from SingHealth. The collaboration can be done through the form of Reverse Pitches, where innovators from SingHealth present clinical needs to a network of start-ups to come up with solutions.
29. Finally, the partnership will develop talent through real-world training opportunities and exchanges. This includes tapping on venture capitalists and multinational companies, to provide on-the-job training opportunities to help more local talent develop innovation and enterprise capabilities.
Closing
30. This partnership will combine the unique strengths of SingHealth and SGInnovate, to harness the vast potential in technology and innovation. And indeed, we are at the cusp of a revolution in healthcare technology.
31. But I like to end my speech today with one reminder, one which all the clinicians amongst us instinctively know. Although I am not one, I know that you will feel so instinctively. Which is, despite the numerous intelligences that technology can mimic, it can never replicate the love, care and empathy of a good doctor.
32. When AlphaGo defeated the world champion in Go, the champion felt despair. AlphaGo felt neither happiness from winning and no desire to cheer. AlphaGo doesn’t love nor hate the game.
33. Healthcare is an industry with high compassion and empathy. In our quest to raise capability and deliver better care, let's make sure AI works alongside us, not as artificial intelligence, but as augmented intelligence that empowers our healthcare workers to have a greater capacity to bring more love, care and compassion to our patients.
34. I wish SingHealth and SGInnovate much success in this important endeavour.
Thank you.